⌕Is ___ SOTA?
isitsota / models / DeepSeek / deepseek-v4-flash-0731
VERIFIEDOct 132026
Is DeepSeek V4 Flash 0731 state of the art?
No.Best recorded rank is #28 in Abstraction (ARC).

As of Oct 13, 2026, DeepSeek V4 Flash 0731 from DeepSeek. Best recorded rank is #28 in Abstraction (ARC). Last listed at #28 in Abstraction (ARC) on 2026-07-31. That board does not award a crown.

Computed automaticallyas of Oct 13, 2026methodology v1.0confidence: low
ReasoningNot SOTA · #29#29 · 0dScienceNot SOTA · #29#29 · 0dFactualityNot SOTA · #57#57 · 0dAbstraction (ARC)Not SOTA · #28#28 · 0d
Alert me if this changesGet badgeCompare with….mdJSON
View as of point-in-time crowns
Why

The ruling, in plain English

Last listed at #29 in Reasoning on 2026-07-31.

Receipts

Evidence

Scores come from the latest ingested snapshot
SourceScoreDateTypeReceipt
GPQA Diamond
deepseek-v4-flash-0731_max · max
0.9103535353535354 0.9–0.92026-07-31INDEP.snap_e03cddec1fbb
SimpleQA Verified
deepseek-v4-flash-0731_max · max
0.33633633633633636 0.3–0.42026-07-31INDEP.snap_e03cddec1fbb
ARC-AGI
deepseek-v4-flash-0731_max · max
0.892026-07-31INDEP.snap_e03cddec1fbb
Reign history

Where it sits in the succession

Full timeline →

No closed or open crown interval includes this model.

Honest about the gaps

Where it is not the crown

CategoryThis modelWhy
ReasoningNot SOTA · #29Last listed at #29 in Reasoning on 2026-07-31.
ScienceNot SOTA · #29Last listed at #29 in Science on 2026-07-31. That board does not award a crown.
FactualityNot SOTA · #57Last listed at #57 in Factuality on 2026-07-31.
Abstraction (ARC)Not SOTA · #28Last listed at #28 in Abstraction (ARC) on 2026-07-31. That board does not award a crown.